In a large distributed computing system of a computing resource service provider, various customers, users, services, and resources of the computing resource service provider are in frequent communication with each other. Keeping the system secure becomes more challenging as applications are updated and as the complexity and usage of the system increases. In such distributed computing systems, it can be difficult to identify the vulnerabilities of the system. Furthermore, it can also be difficult to isolate and troubleshoot issues with the system, and even when issues are identified, reliably securing the system may be challenging. Additionally, many conventional systems rely upon manual mitigation of such vulnerabilities. The complexity and distribution of computing resources in these environments may make it difficult to collect and analyzes log information generated by the computing resources.